https://retellai.com/?utm_source=discord&utm_medium=seo logo
Powered by
where can I see concurrency charges
# ❓・ai-evy-support
a
@AI Evy where can I see concurrency charges
a
Thanks for tagging me @Alex. I'll look through relevant sources and get back to you shortly.
### Answer You can view concurrency charges in the Usage tab on the Billing page. [(1) ]() The Usage tab provides a Cost by Provider breakdown, which includes concurrency usage among other expenses like voice engines, LLMs, and telephony services. To upgrade your concurrency limit or purchase additional concurrent calls, go to the Billing page in the dashboard. [(2) ]() You can also see your live concurrent call usage in the bottom-left corner of the dashboard. [(3) ]()
​ To learn more, see the sources I considered: [[1] Billing Overview](https://docs.retellai.com/accounts/billing) [[2] Understand concurrency & limits](https://docs.retellai.com/deploy/concurrency) ​
Feel free to tag @AI Evy with additional questions. I also appreciate feedback, it helps me improve.
a
@AI Evy i want to be able to book 20 barbers with my voice AI agent but iw ant a simple build, so i wwas think if i could have a dynaimc variable in the convertsadytion that is the barber name that was desired by the user as the variable to use in the APIU call to get availablility from the valendar, can i use a dnaynmica variable in the APIU call to get the barber calendar that the person wants, instead of having 20 unique barber check avialbilty functions? can i do it simpler?
a
Thanks for the follow-up, @Alex. I'll get back to you soon.
### Answer Yes, you can simplify this by using dynamic variables in your tool configurations. Dynamic variables can be used in custom function URLs and tool descriptions. [(1) ]() Here's the approach: 1. Collect the barber name during the conversation and store it in a dynamic variable (e.g., 
{{barber_name}}
) . 2. Use that dynamic variable in a single check availability function — you can include 
{{barber_name}}
in the custom function URL or pass it as a parameter to your API endpoint. This way, you only need one check availability function that dynamically references whichever barber the user selected, rather than creating 20 separate functions. For the calendar integration itself, you'd configure a check calendar availability function with: - Tool name (unique within the agent) - API Key from cal.com - Event Type ID - Tool description [(2) ]() The key is that your backend API would receive the barber name as a parameter and route to the appropriate calendar based on that value. You can also use collected dynamic variables from the call to pass user-provided information (like the barber name) to your function calls. [(11) ]()
​ To learn more, see the sources I considered: [[1] Dynamic Variables](https://docs.retellai.com/build/dynamic-variables) [[2] Check Calendar Availability](https://docs.retellai.com/build/check-availability) ​
Feel free to tag @AI Evy with additional questions. I also appreciate feedback, it helps me improve.
a
@AI Evy build me a custom payload for getting avialability of the barber with that dynamic variable for cal.com
a
Thanks for the follow-up, @Alex. I'll get back to you soon.
### Answer --- Here's an example custom function parameter schema that uses a dynamic variable for the barber name:
Copy code
json
{
  "type": "object",
  "required": [
    "startDate",
    "endDate"
  ],
  "properties": {
    "barber_name": {
      "type": "string",
      "const": "{{barber_name}}"
    },
    "startDate": {
      "type": "string",
      "description": "start date in YYYY-MM-DDTHH:MM:SSZ format"
    },
    "endDate": {
      "type": "string",
      "description": "end date in YYYY-MM-DDTHH:MM:SSZ format"
    }
  }
}
[(1) ]() , [(2) ]() , [(3) ]() Key points: - Use 
"const": "{{barber_name}}"
to pass the dynamic variable directly to your API — const values are applied directly to the function without LLM resolution. [(1) ]() , [(2) ]() - The 
startDate
and 
endDate
use 
"description"
so the LLM resolves them based on what the user says. [(1) ]() , [(2) ]() - Make sure your date format strictly matches 
YYYY-MM-DDTHH:MM:SSZ
with the 
Z
indicating UTC time. [(3) ]() Your backend API would then receive the 
barber_name
parameter and use it to query the correct cal.com calendar/event type for that barber. You can also include the dynamic variable in query parameters or the endpoint URL itself if your API structure requires it. [(1) ]() , [(2) ]()
​ To learn more, see the sources I considered: [[1] Integrate any system with custom function](https://docs.retellai.com/build/single-multi-prompt/custom-function) [[2] Custom Function](https://docs.retellai.com/build/conversation-flow/custom-function) [[3] custom function unable to read dynamic variables as payload's input](https://discord.com/channels/1186424376300679209/1351866528680902697) ​
Feel free to tag @AI Evy with additional questions. I also appreciate feedback, it helps me improve.
a
@AI Evy i have a request to create a japanese version of one of the agents that we built on retell, a particular request is that her accent and colloquial language match that of the region of japan, can you fgive me instructions for my chatgpt agent that could help guide this process? be as detailed as possible, evven referencing retell specifci fucntionality so that the chat is contextualized exactly to the goal
a
Thanks for the follow-up, @Alex. I'll get back to you soon.
### Answer --- I don't have specific documentation on creating region-specific Japanese accents or colloquial language patterns in my available sources. However, here's what I can share about the relevant Retell configuration options: Language Setting Set the agent's language to 
ja-JP
for Japanese speech recognition. [(1) ]() , [(2) ]() [(1) ]() , [(2) ]() Voice Temperature The 
voice_temperature
parameter controls how stable vs. variant the voice is, ranging from 0-2. [(1) ]() , [(2) ]() [(1) ]() , [(2) ]() Pronunciation Dictionary You can add custom pronunciation entries to guide speech synthesis for consistent pronunciation of specific words/phrases (currently supported for English & 11labs voices) . [(1) ]() , [(2) ]() [(1) ]() , [(2) ]() Prompt Engineering For regional dialect and colloquial language, you'd primarily handle this through your LLM prompt — instructing the agent to respond using specific regional expressions, honorific patterns, or dialect markers (e.g., Kansai-ben, Tohoku dialect, etc.) . --- For detailed guidance on Japanese regional dialects and voice characteristics, I'd recommend: Check out these potentially useful sources: - [Configure LLM Options](
ns) - [Create Voice Agent]() You may also want to explore the available Japanese voices in the Dashboard to find one that best matches your target regional sound. Would you like me to try providing a general answer on Japanese regional dialect prompting strategies?
​ To learn more, see the sources I considered: [[1] Create Voice Agent](https://docs.retellai.com/api-references/create-agent) [[2] List Voice Agents](https://docs.retellai.com/api-references/list-agents) ​
Feel free to tag @AI Evy with additional questions. I also appreciate feedback, it helps me improve.
a
@AI Evy yes build me comprhesnive plan to the best of your knowledge
a
Thanks for the follow-up, @Alex. I'll get back to you soon.
### Answer ### Answer I don't have detailed documentation on Japanese regional dialects or accent engineering in my available sources. However, I can provide a comprehensive plan using the Retell-specific configuration options I do have documented: Step 1: Agent Language Configuration Set the language to Japanese: [(1) ]()
Copy code
json
"language": "ja-JP"
Step 2: Voice Selection & Tuning Configure voice parameters:
Copy code
json
"voice_temperature": 1,
"voice_speed": 1
- 
voice_temperature
ranges from 0-2; higher values create more variant speech generation (may help with natural-sounding regional inflections) - 
voice_speed
ranges from 0.5-2; adjust to match the natural speaking pace of the target region Step 3: Speech Normalization Enable speech normalization for consistent pronunciation of numbers, dates, and currency:
Copy code
json
"normalize_for_speech": true
*Step 4: Prompt Engineering (LLM-Side) * For regional dialect and colloquial language, your LLM prompt should include: - Specific regional expressions and vocabulary - Appropriate honorific patterns (keigo levels) - Regional dialect markers (e.g., Kansai-ben endings, Tohoku pronunciation notes) - Character personality traits that align with regional stereotypes if desired Step 5: Boosted Keywords Add region-specific vocabulary to improve transcription accuracy:
Copy code
json
"boosted_keywords": ["regional_term_1", "regional_term_2"]
Limitations: - Pronunciation dictionary is currently only supported for English & 11labs voices For the actual dialect content and linguistic guidance, you'd need to work with a Japanese language specialist or use external resources, as this falls outside Retell's technical documentation. Check out these potentially useful sources: - [Create Voice Agent]() - [Update Voice Agent]() Would you like a general answer on Japanese regional dialect prompting strategies based on broader knowledge?
​ To learn more, see the sources I considered: [[1] Create Voice Agent](https://docs.retellai.com/api-references/create-agent) ​
Feel free to tag @AI Evy with additional questions. I also appreciate feedback, it helps me improve.
a
@AI Evy what is system time for tokyo japan
a
Thanks for the follow-up, @Alex. I'll get back to you soon.
### Answer To get the current time for Tokyo, Japan in your Retell agent, use the timezone-specific dynamic variable: [(1) ]()
Copy code
{{current_time_Asia/Tokyo}}
Retell supports system default dynamic variables including 
{{current_time}}
for America/Los_Angeles, and 
{{current_time_[timezone]}}
for any given timezone. [(2) ]() Example timezone formats: - 
{{current_time_Asia/Tokyo}}
— for Tokyo, Japan - 
{{current_time_America/New_York}}
— for Eastern Time - 
{{current_time_Australia/Sydney}}
— for Sydney , [(1) ]() The format returns the current time like: 
Monday, March 24, 2025 at 02:53:01 PM PDT
[(2) ]() You can add this variable to your prompt or begin message to give your Japanese agent accurate local time context.
​ To learn more, see the sources I considered: [[1] Dynamic Variables](https://docs.retellai.com/build/dynamic-variables) [[2] Platform Changelogs | Retell AI](https://www.retellai.com/changelog) ​
Feel free to tag @AI Evy with additional questions. I also appreciate feedback, it helps me improve.
Thanks for the follow-up, @Alex. I'll get back to you soon.
a
@AI Evy why are call transfers failing when we try to transfer from 1 agent phone number to another
a
### Answer There are several common reasons call transfers fail: 1. Call Type Call transfer is only supported for phone calls, not web calls. [(1) ]() 2. Telephony Issues Call transfer is similar to placing an outbound call, and failures occur for similar reasons as outbound call failures. The SIP connection log is available in the call logs and is useful for diagnosing the failure reason. 3. Human Detection If human detection is enabled, the call may fail if a human is not successfully detected. Possible reasons include: - No human was actually present (e.g., it was an IVR or voicemail) - The other party spoke, but only after the detection timeout expired - The speech was too similar to an IVR or voicemail and was not recognized as human 4. SIP Inbound Limitation Based on community reports, SIP inbound calls may not transfer successfully. [(2) ]() 5. Opt-Out Option Conflict One user reported that having the opt-out option activated caused transfer failures — deactivating it resolved the issue. [(3) ]() Troubleshooting Steps: - Check your call logs for SIP connection details and error messages - Verify the transfer destination number is correct and reachable - Confirm you're testing on actual phone calls, not web calls - Review your transfer function configuration to ensure it's properly set up [(1) ]() If you're seeing 
{"status":"failed to transfer"}
, this can also occur when the destination simply doesn't pick up. [(4) ]() Can you share any error messages from your call logs or the specific tran
sfer configuration you're using?
​ To learn more, see the sources I considered: [[1] Debug call transfer failure](https://docs.retellai.com/reliability/call-performance) [[2] All transfers failling](https://discord.com/channels/1186424376300679209/1417958705957310555) [[3] Discord message from @_sulieman on 2024-12-12](https://discord.com/channels/1186424376300679209/1294891368505937952/1316854187958730794) [[4] Failed transfers on all calls](https://discord.com/channels/1186424376300679209/1297930701009588246) ​
Feel free to tag @AI Evy with additional questions. I also appreciate feedback, it helps me improve.
a
@AI Evy this was what we got from gthe logs, we are just using retell numbers: Agent: Transferring you to our Japanese agent now. 0:13 Tool Invocation: transfer_call tool_call_id: transfer_call_10677e { "number": "+1(647)584-0591", "option": { "type": "cold_transfer", "showTransfereeAsCaller": true, "enableBridgeAudioCue": true } } Tool Result tool_call_id: transfer_call_10677e {"status":"Transfer call cannot be completed, telephony provider or carrier declined or was unavailable.. Please inform the customer that the transfer did not go through and offer to try again or assist them directly."}
a
Thanks for the follow-up, @Alex. I'll get back to you soon.
### Answer The error 
"telephony provider or carrier declined or was unavailable"
indicates an issue at the telephony/carrier level during the transfer attempt. [(1) ]() Issue with your transfer number format: Your transfer destination is formatted as 
"+1(647) 584-0591"
— this contains parentheses which is not valid E.164 format. [(2) ]() The transfer number must be in e.164 format (e.g., 
+16475840591
) or a SIP URI in the format 
sip:username@domain
. Fix: Change your transfer destination from:
Copy code
+1(647) 584-0591
to:
Copy code
+16475840591
Additional context: Call transfer failures occur for similar reasons as outbound call failures. [(1) ]() The 
telephony_provider_unavailable
disconnection reason indicates the telephony provider is unavailable. [(3) ]() One community user experiencing similar transfer failures with a custom telephony provider resolved the issue by creating a new SIP trunk connection with their PBX. [(4) ]() Since you're using Retell numbers, ensure the destination number is correctly formatted and reachable. If the issue persists after fixing the format, check the SIP connection log in your call logs for more detailed error information. [(1) ]()
Feel free to tag @AI Evy with additional questions. I also appreciate feedback, it helps me improve.
a
@AI Evy give me system time for toronto eastern time
a
Thanks for the follow-up, @Alex. I'll get back to you soon.
### Answer --- To get the current time for Toronto (Eastern Time) in your Retell agent, use the timezone-specific dynamic variable: [(1) ]()
Copy code
{{current_time_America/Toronto}}
Retell supports 
{{current_time_[timezone]}}
for any specified timezone. The output format will be similar to: 
"Thursday, March 28, 2024 at 11:46:04 PM EST"
You can include this variable in your prompts or begin message to give your agent accurate local time context.
​ To learn more, see the sources I considered: [[1] Dynamic Variables](https://docs.retellai.com/build/dynamic-variables) ​
Feel free to tag @AI Evy with additional questions. I also appreciate feedback, it helps me improve.
a
@AI Evy is there any limits to knowledge base?
a
Thanks for the follow-up, @Alex. I'll get back to you soon.
### Answer Yes, there are several limits for knowledge bases: [(1) ]() Per Knowledge Base Limits: - URL: Maximum 500 URLs - Auto-Crawling URL Paths: Maximum 200 exclusion URLs for each auto-crawling path, and maximum 500 exclusion URLs per knowledge base - Text: Maximum 50 text snippets - File: Maximum 25 files, with each file limited to 50MB. For CSV, TSV, XLS, and XLSX files, the row limit is 1,000 rows and the column limit is 50 columns Workspace Limits: - First 10 knowledge bases are free per workspace , [(2) ]() - Additional knowledge bases are billed at $8/month per knowledge base [(1) ]() You can create multiple knowledge bases to overcome these limits, and an agent can have more than one knowledge base linked to it. Pricing for Usage: - $0.005 per minute of calls that have knowledge base enabled (regardless of how many knowledge bases are linked)
​ To learn more, see the sources I considered: [[1] Knowledge Base](https://docs.retellai.com/build/knowledge-base) [[2] Simplify Support with Retell AI’s Knowledge Base | Retell AI](https://www.retellai.com/blog/knowledge-base-feature) ​
Feel free to tag @AI Evy with additional questions. I also appreciate feedback, it helps me improve.
a
@AI Evy so if you have alot of knowledge bases how does the agent know which one to retreive from?
a
Thanks for the follow-up, @Alex. I'll get back to you soon.
### Answer ### Answer When you have multiple knowledge bases linked to an agent, the retrieval works automatically based on relevance. [(1) ]() During a call, when the agent is about to respond, it uses the transcript so far (prompt is not included) to find the most relevant chunks from the knowledge base and feeds them to the LLM as context. The agent doesn't "choose" a specific knowledge base — it searches across all linked knowledge bases and retrieves the most relevant chunks based on semantic similarity to the conversation. Configuration options: You can adjust retrieval behavior via "Adjust KB Retrieval Chunks and Similarity": - Chunks to retrieve: Max number of chunks to retrieve (range 1-10, default: 3) - Similarity Threshold: How strict matching is — higher values give fewer but more similar matches (default: 0.6) For Conversation Flow agents: You can also configure knowledge bases at the node level, which combines with the agent's global knowledge base to provide more accurate context for specific topics. Important note: The knowledge base name does not influence retrieval — it's for display purposes only and is not included in the retrieval process.
​ To learn more, see the sources I considered: [[1] Knowledge Base](https://docs.retellai.com/build/knowledge-base) ​
Feel free to tag @AI Evy with additional questions. I also appreciate feedback, it helps me improve.
a
@AI Evy does it matter what we name the knowedlge bases?
a
Thanks for the follow-up, @Alex. I'll get back to you soon.
### Answer No, the knowledge base name does not influence retrieval. [(1) ]() The name is for display purposes only and is not included in the retrieval process.
​ To learn more, see the sources I considered: [[1] Knowledge Base](https://docs.retellai.com/build/knowledge-base) ​
Feel free to tag @AI Evy with additional questions. I also appreciate feedback, it helps me improve.
a
@AI Evy does tabbing and spacing matter in JSON body for KB?
a
Thanks for the follow-up, @Alex. I'll get back to you soon.
### Answer --- I don't have specific documentation on whether tabbing and spacing affects JSON body parsing for knowledge base API requests. However, standard JSON formatting rules apply — JSON is whitespace-insensitive, meaning tabs, spaces, and newlines between elements are ignored during parsing. What matters is valid JSON syntax (proper quotes, commas, brackets, etc.) . For the knowledge base API, the body format uses 
multipart/form-data
for requests that include files, and the JSON fields like 
knowledge_base_texts
and 
knowledge_base_urls
should follow standard JSON syntax. [(1) ]() [(2) ]() Check out these potentially useful sources: - [Create Knowledge Base]() - [Add Knowledge Base Sources]() Would you like a general answer on JSON formatting best practices?
​ To learn more, see the sources I considered: [[1] Create Knowledge Base](https://docs.retellai.com/api-references/create-knowledge-base) [[2] Add Knowledge Base Sources](https://docs.retellai.com/api-references/add-knowledge-base-sources) ​
Feel free to tag @AI Evy with additional questions. I also appreciate feedback, it helps me improve.
a
@AI Evy wy cant i change knowledge base name?
a
Thanks for the follow-up, @Alex. I'll get back to you soon.
### Answer I don't have documentation on how to change a knowledge base name or whether there are restrictions preventing name changes after creation. The documentation shows that 
knowledge_base_name
must be less than 40 characters [(1) ]() [(2) ]() , but I don't see information about editing or updating the name after the knowledge base is created. Check out these potentially useful sources: - [Knowledge Base]() - [Get Knowledge Base]() You may want to check if there's an Update Knowledge Base API endpoint, or reach out to support for clarification on this functionality.
​ To learn more, see the sources I considered: [[1] Get Knowledge Base](https://docs.retellai.com/api-references/get-knowledge-base) [[2] List Knowledge Bases](https://docs.retellai.com/api-references/list-knowledge-bases) ​
Feel free to tag @AI Evy with additional questions. I also appreciate feedback, it helps me improve.
PreviousNext
Powered by